Each year, over 250,000 hip and 500,000 vertebral fractures occur in the U.S., with associated health care costs in the billions of dollars. While hip fractures are predominantly traumatic events, vertebral fractures are largely progressive, and almost one in three women aged over 65 suffer from this pathology. The strength of vertebral bodies is dominated by the strength of the trabecular bone within its cortices. However, while most biomechanical studies of vertebral bodies have focused on disc degeneration and bone fragility as determinant factors in fracture etiology, few studies have addressed how stresses or strains within the vertebral body are sensitive to loading conditions or to localized damage of trabecular bone. Recent pilot experiments have revealed that the post-yield (loading to non-linear regime, unloading, and reloading) behavior of trabecular bone is characterized by permanent reductions in modulus and strength on reloading, behavior which is typical of damaged materials. Simple strength of material analyses indicate that post-yield behavior may be important in failure processes of whole bones because of subsequent stress protection of damaged bone, stress redistribution to undamaged bone, and consequent failure of previously undamaged bone. Finite element predictions indicate that stresses in the cortical shell and end plates of vertebral bodies may be very sensitive to the structural integrity of the trabecular bone. These findings suggest that whole bone failure may be sensitive to trabecular bone post-yield behavior. Thus better understanding of the structural consequences of post-yield behavior of trabecular bone may lead to identification of loading regimes which compromise the structural integrity of whole bones. The overall goal of this research is to test the hypothesis that whole bone failure etiology can be highly dependent on load history, especially on loads which cause localized damage of trabecular bone, resulting in substantial stress redistribution to, and failure of, previously undamaged bone. We will approach this problem in a staged fashion, starting with a relatively simple structure and then progressing to study more complex structures which better simulate the in vivo situation. In particular, we will experimentally formulate a constitutive law for trabecular bone post- yield behavior, and develop methods to quantify damage in embedded specimens. Next, we will incorporate this behavior into experimentally verified finite element models. Finally, using detailed finite element analysis of whole vertebral bodies, which will be rigorously verified by a series of in vitro experiments, we will study consequences of post- yield behavior for vertebral fracture etiology.